Surveylab has come a long way since humble beginnings in April, 2003. I remember the early days well.

John was offering consulting services under his CRL Solutions hat while I developed the online survey and reporting platform. Our very first client, Gatwick Airport/BAA, has returned to us every year since to conduct at least one more survey… John introduced me to an old boss of his in a cafe under Charing Cross which was to be the start of a partnership with our second client who became good friends too. That June, I fielded our second survey (for Colin) in 10 countries including Japan and Mexico.

After that it’s a bit of a blur, apart from installing our first servers at an exemplary web hosting company called UK Dedicated in August, 2003…

We didn’t make enough in the first year to pay either John or myself a proper salary. Truth be told, after about 15 months Surveylab was perilously close to running out of funds because the plan was always to generate the income on which to grow.

When I look back now I marvel at how I used to feel so busy and under pressure juggling just 2 or 3 different clients’ surveys, or how we would focus on things that seemed important but turned out to be irrelevant (including learning to identify when we were “quote fodder”). In the beginning there was no plan to hire many staff at all – we were going to develop D.I.Y. survey software which would look after itself! – but our few clients liked having the expert feedback and a professional edge combined with flexibility. We had found our niche, although it took us some time to realise it.

Surveylab has grown each year in a number of ways: revenue (obviously) and staff (now 8), number of clients (estimated close to 200) and types of survey too. Coming from TARP, our background was “customer satisfaction measurement”, but in January, 2005 we ran our first employee engagement survey. Since then, we learned how far people walk to their bank to deposit cheques, the average payment made by the Tooth Fairy, and inadvertently supported the winners of Euro 2008 on the back of many, many P.R. led surveys.

And all the time, the number of customer feedback surveys has continued to grow (with special thanks to our friends at CCMC in Washington).

Each year Surveylab’s birthday passes more or less unheralded. We are nearing a big milestone though and in the last eighteen months I’ve felt a change; our products are mature, we are a great team and I am comfortable with the term experts (but not guru, please), I feel justified in using the word experienced.

So yesterday, Hew and I were testing Iftikhar’s latest code that makes the task of creating a survey in multiple languages so much easier. Charles oversaw the sending of 16,530 emails for half a dozen surveys in between working on the next surveys that will launch next week. Georgiee was working on marketing material (in the realms of patient engagement) and Cathy received her mail, and began processing/entering the next batch of printed questionnaires… I’m not sure what John was up to but I bet it involved talking to clients.

Our birthday (23rd April) passed like any other day. Perhaps we’ll have a party next year!

However, the email invitation doesn’t always get quite the attention it deserves yet it’s the first thing your customers, staff or potential respondents will see. Here are ten tips to check on your next survey email that can help increase your survey’s response rate.

1. Personalise the Survey Invitation
   Emails are more likely to be opened and read if the message starts with greeting your customer/employee/recipient by name. There are academic studies that prove this, and email hosts (like Hotmail and Gmail) use the absence of a personalised greeting to help decide whether a message may be spam (or simply not unwanted). When we receive contact emails from a client without names, we immediately lower our expectation for response rate.
2. Include Only One Call-To-Action
   Keep the message simple! Don’t offer additional links to promote this or that – ultimately, your goal is to collect feedback so your ‘Call To Action’ (CTA) should be something like click here to take the survey and a link that takes the user to the survey.
3. Use Current, Accurate Customer Lists
   Although it may be tempting to send a survey to every last customer you have an email for – consider carefully which customers to include in your mailing. Is there really value in asking for feedback from customers who have not purchased or contacted your company in the last 18 months? You need reliable data – think quality, not quantity.Those customers who purchased/contacted recently will be far more likely to complete a survey (and recall their experience) than those from a few months or longer ago.
4. Send a Reminder
   This one action will boost the number of responses received so far – often we see the numbers double, and regularly see at least an additional 50% to the completed surveys so far.
   
   The reminder message works best if it is targetted to the non-respondents only (in other words excluse those who have already completed the survey) because then your message can be very focussed and doesn’t need to explain “if you have completed the survey already, thank you” (again, think one CTA).
   
5. Provide Contact Details
   Including your organisation’s address adds credibility that the survey is genuine and a degree of professionalism. In the U.S. the CAN-SPAM Act requires a physical address to be present in the email, and email hosts are purported to check for their presence before accepting the message through.
6. Use your Branding
   Take a look at the newsletters, order confirmation and customer support emails that your organisation sends out. If people are used to receiving email from ‘Your-OrganisationSupport’ then try adapting this to use in your survey email too.There is no right answer on what works, the important thing here is to test. Experiment with the Branding in the From field, or in the Subject line, and the content of your email to find what works best for your survey.
7. Use and Monitor the Reply-To Address
   If you send 1,000 emails, and receive 300 out-of-office replies (it happens), this can help explain activity and you may adjust your schedule or plans for reminder emails. Similarly, if you receive 200 ‘bounces’ (undeliverable emails because the address is wrong/no longer current) then your response rate is better than you thought (do feed this information back to who-ever manages/created the email list).For most surveys, a few people will email back because they have a question or problem. This feedback may be useful to communicate to other participants (e.g. update the FAQs, opening or reminder message) or in extreme cases highlight an issue with the survey itself. With customer surveys, a handful of replies inevitably ask for product support/service – which may be an opportunity to make the survey pay for itself in other ways by delivering service recovery.
8. Advanced Communications (for Employee Surveys)
   Arrange for an email or communication to come from the CEO before the survey invites are sent out but, hopefully this shouldn’t be the first that staff hear about their employee survey – communication is key!
9. Set Customers Expectations about Feedback
   Have you noticed how shop staff don’t seem to insist on stapling your receipts together any more, instead they grab a highlighter marker and point you to a survey they’d like you to take? When capturing customer contact details, state that one of the purposes of collecting email is that you conduct feedback surveys. In the order confirmation email, include a statement about leaving feedback in a future survey (but contact Customer Services if you have a question or problem).
10. Never Use BCC to Send Survey Invites

Do you have any tips to share? Post a comment below, or send me a tweet!

Related posts: A few examples of invite messages are posted under the email side of surveys.

In April last year, we analysed the traffic logs for our customer-facing surveys* and back then learned that across most surveys just 1% of users clicking through were on a mobile device. Less than 12 months later, the number of mobile users has jumped to 8%!

Most surveys are still completed on a Windows PC, but use of mobile devices jumped to 8% by January 2012

* We analysed traffic logs of customer surveys conducted in North America from October 2011 to mid-January 2012. The surveys were a mix of B2B and consumer-targeted surveys. We only looked at the traffic stats for customer surveys because we expect employee surveys to have a disproportionate number of deskbound users.

Looking at the individual mobile devices, iPhone users accounted for 3.5% of our survey respondents:

Is this important?

I think so. Online surveys are rarely (intentionally) designed for anything smaller than a 12-inch screen. The assumption is that the respondent is sitting in front of a PC or laptop when they access the survey. Various reports claim between 40-50% of mobile-phone users own smart phones and mobile devices, and while I don’t expect the desktop PC to disappear anytime soon, people’s surfing habits and where they access their emails is changing.

Microsoft's Internet Explorer still dominates, but do you ever check how the survey looks or behaves in any other browser or screen-size?

Our findings raised a fair few questions like

• Do mobile users still complete the survey?
• Will mobile users complete a survey that takes a desktop user 10 minutes to complete?
• Why do mobile users bail out of the survey? Their connection, timing, etc.

I don’t have the answers yet.

If you have access to analytics data for your website, look at the stats for mobile users, then take another look at your survey from their perspective.

You may also like: Is your survey compatible with smartphones and iPads?

I’ve talked before about how to make the task a little easier, starting with a word cloud to look for common words and themes. You might also try reviewing the comments broken down by demographics – for example going through an employee survey’s comments received by the Finance Department, then I.T, or perhaps by depots/stores/regions. A little forward planning also helps – how you ask for the qualitative feedback in your survey can have a big effect on what people write.

Last month, we introduced our latest new reporting feature ‘Verbatim Tagging’ – and it is proving extremely useful in analysing survey comments.

What is Verbatim Tagging?

Each comment (or verbatim) can be tagged with whatever classification or category you choose. Some research analysts would call it coding/recoding answers – as you read through the comments you tag those that are about “poor quality”, “delivery delay”, “communications” or “pay & rewards” etc. When you are done, you can more quickly refer to and find specific comments on a topic. As a handy consequence, you also end up with a count of how many comments are tagged with each category.

That’s all verbatim tagging is – categorising the qualitative comments received. You may have tried doing something similar in Microsoft Excel but reading lots of text in an Excel worksheet is not that nice, and once you have more then 5 or 6 columns (categories), it’s easy to screw-up too. When we analyse comments we used to create and apply filters to the results, but tagging is much, much easier to use.

How to use Verbatim Tagging

Current Surveylab clients should have Tagging available now. Simply log into your online reports, go to the Results tab and view the comments received for any verbatim question.

Screenshot: Tagging a survey comment in Surveylab's online reports

To tag a comment, simply click on the Tags link, and type your tag. Use a comma to apply multiple tags.

All tags are shown at the top of the results page (the number next to the tag shows how many comments have this tag). Click on the tag at the top of the page to view only those comments.

Screenshot: Viewing the survey comments and their tags

Tips!

• If you accidentally insert a typo into a tag – click on the comment’s tag and change it.
• View comments by a particular tag, then select which comments you want to filter by in the normal way. If you want to create a filter that spans 2 or more tags: click on the tag to view, then tick the comments to filter (but don’t click ‘Apply Filter’). Click the alternate tag to view at the top of the page and now select which of these comments to filter. Now click ‘Apply Filter’ – all comments selected (ticked) are part of your filter.

If you would like a quick-run through, or a full reporting demo, please ask. We’re happy to show off (contact info here).

A lot of the content from the guide has been added to and published on our blog. Below is a quick overview of those posts:

1. Getting started with Employee Surveys
   A summary of the survey process from start to finish. First task: Define the purpose
2. Hearing and believing: Good employee surveys need GREAT communication
   Promote the survey! Planning communications throughout the study, and in advance, is essential for success
3. Choosing to print your questionnaire or email the survey
   Benefits and costs of online versus printed surveys
4. Employee Surveys – to outsource or manage in-house?
   Expertise, technical capability, resources, experience and speed… and improved anonymity
5. Essential survey questions: Willingness to recommend
   “Would you recommend” is an insightful question to include in most surveys
6. Choosing questions to ask in an employee survey
   Looking at more questions and themes in employee surveys
7. Survey scales – can I use a 3/5/7/10 point scale?
   Choose a range of answers that works for you
8. Using verbatim comments in an employee survey
   Tips on asking (and analysing) open-ended questions in a survey
9. A checklist for testing your survey before you launch
   Things to test and look out for before your let your survey loose!
10. Your survey has launched – what next?
    Monitor your survey and email inbox in case staff ask for help
11. How to analyse your employee survey’s results
    An overview of making sense of the results and what to do with them

If you would like a PDF copy of our guide (A How-To Guide for Better Employee Surveys) from which some of the articles above are drawn please email me – hello@surveylab.co.uk. We are happy to help!

Want to read more?!

Use the ‘Filter Posts by Category’ links on the right hand side of the page to browse all posts under a topic, e.g. Employee Surveys. Other popular posts that may be of interest are:

• How much does an employee survey cost?
• How do you link employee engagement to customer satisfaction?
• What is an average survey response rate?

We welcome any feedback or feel free to ask questions: email us, post a comment or follow me on Twitter!

A while ago, I came across some excellent (and simple) guidelines that can help managers identify whether changes in their survey results over time are important or not. We often use similar approaches when we help a client make sense of their results, but Jack Wiley, the author of Strategic Employee Surveys, describes his method clearly with good examples to explain. His guidelines are (my words):

If the unit (e.g. team/department/region) has less than 50 responses an increase or decrease of 15% or more is a “practically” significant change.

If the unit has between fifty and 100 responses – a change of 10% is significant.

More than 100 responses – the change is significant if the results have increased or decreased by 5%.

For example, in a team of 20 people, each person’s score contributes a minimum of 5% to the score that will be reported for the team. This means that just 3 or 4 people’s answers can affect the score noticeably, so the smaller the group, the larger the difference required (which is a basic rule of statistics).

Mr Wiley does mention a couple of limitations (these guidelines are for small organisations/units and aimed at comparing a question’s scores over time) – but hopefully this tip helps the odd H.R. Manager or head of department recognise what some of the numbers mean!

I highly recommend the book Strategic Employee Surveys: Evidence-Based Guidelines for Driving Organizational Success – also reviewed in the CIPD magazine (sign-up required).

The responses are collated, the data aggregated and results churned out. Some proper analysis follows, reports produced, culminating in discussions and investigations, changes and improvements. Powerful stuff!

If this happens at your organisation – well done.

Customers are individuals too

But what happens to the dissatisfied customer(s) in the process above? There is an assumption that any issue the customer has will already have been reported (and resolved) or is being dealt with via official channels. However, for the customer the survey is another way to try and engage the organisation to address their issue(s). Make the most of this opportunity! In addition to providing aggregated data that will create insight and actionable findings, a customer survey can pay for itself by helping perform ‘service recovery’ amongst dissatisfied customers.

This is why so many people love Twitter – you can shout beetlejuice three times and as if by magic the shopkeeper appears.

Using a customer survey to perform ‘service recovery’

It should be relatively easy to tweak a customer survey by adding the question

Would you like [name of organisation] to personally contact you about your feedback?

When the survey is completed, an email can then be triggered or details appended to a daily report that will be reviewed and acted upon by an escalation team. Turning disgruntled customers into satisfied customers can increase loyalty substantially resulting in repeat business and referrals that can be worth a lot more than the cost of conducting the customer survey.

Identifying ‘Customers at Risk’

The survey process may also be modified to identify what we call ‘Customers at Risk’. These customers may not ask to be personally contacted (or you may not offer the option in the survey), but if a response meets pre-defined criteria (for example, willingness to recommend scores 6 or less on a 10 point scale) then these customers will be tagged as at risk. Again, this information should be passed to an escalation team for review and remedial action.

If you do use service recovery, remember you cannot claim the survey is anonymous and strictly confidential to respondents. Consideration should also be made for how customer at risk information is shared within the organisation.

Service recovery is most successful when used in a ‘transactional survey’ (i.e. a survey to customers who recently completed a ‘transaction’ be it a purchase or other form of contact). The survey needs to follow as soon as possible after the transaction so that the experience is still fresh in the customer’s mind and also close enough to the event that feedback remains actionable and relevant. It may not be possible or practical to conduct your surveys regularly (monthly) but aim for as timely as you can.

A well handled problem builds more loyalty than no problem at all

Occasionally, a survey detractor will argue that the organisation asks the questions but doesn’t listen to the customer. However, a good customer survey can deliver more than insight and answers; it can directly lead to more loyal customers!

You may also like:

1. Getting Customer Feedback: Do it right or don’t do it at all
2. Five Tips for a better customer survey in 2012

Planning a customer survey in 2012? Surveylab has years of experience designing, implementing and managing customer feedback programmes (and more from past careers and partnering with specialist consultancies). Contact us for an informal chat to discuss your plans. We’re happy to help.

A well designed customer feedback survey will deliver actionable data. If the survey is used simply to tell your staff how good you have been, and nothing more is done with the results, then yes, the survey has not lived up to its potential.

For example, a carefully designed customer survey can tell an organisation not only when something is a recurring problem, but also how much that problem costs in terms of lost business or the subsequent support, which helps prioritise follow-up actions.

And this is the key: follow-up action.

The survey is a tool. It provides results that need to be analysed to deliver answers/insight/information/findings on which the organisation can act. Whether you use Net Promoter or ServQual, a five point scale (or four or seven) is not important. It’s what you do with the results afterwards that counts.

Social media is the same but currently is drawing a lot of attention and effort into using it effectively. Twitter and facebook are excellent tools to engage with customers and collect feedback, but I don’t see them as a replacement to the survey… just yet.

Need help designing/implementing a customer feedback survey? Talk to Surveylab.

1. Are you measuring what’s important?

Does your survey generate actionable data, or is it a scorecard that simply allows you to pat yourself on the back when the results are published?

If your survey wasn’t reviewed during 2011, re-evaluate your survey in early 2012. We often hear a requirement that “we want to keep this and that because we want to compare to last year” but are these items still relevant and important today? Key drivers of customer loyalty may need revising, especially if your organisation has undergone significant change in recent years.

2. Keep the survey short

I say this all the time. Customers will complete surveys, but they are busy, and online surveys have a lot more to compete with for attention today than five years ago. A prize-draw incentive doesn’t often make a lot of difference; the customer survey is a chance to give feedback in the hope of improved or continued good products/services in future – and this shouldn’t be taking more than ten minutes to complete.

3. Run surveys more frequently

Some customer research is better conducted as a ‘transactional survey’ because customers will not remember details over time. For example, feedback on complaint handling or product repair should be requested within days/week of the contact, rather than presented as another section of an annual customer survey. Why wait 12 months for an update on progress?

4. Use surveys to perform service recovery

A survey can pay for itself by helping service recovery amongst unhappy customers. Customers often want information and for their voice to be heard, not compensation (PDF link), and adding email alerts to notify relevant teams about dissatisfied customers is easy to do. ‘Customer at Risk’ reports help identify recurring problems too, and when combined with running surveys more frequently it is a lot easier to convert the disgruntled customer into a satisfied customer (or perhaps a raving fan!).

5. Conduct proper analysis and act on findings

What happens after the survey reports are published? Who gets the feedback?

Don’t restrict results to senior management – local managers and front line staff can benefit from, even be inspired by, customer insights too (although keep it relevant). Your staff may even have recommendations(!) and can take action from the ground up, rather than wait for direction to filter down.

Planning a customer survey in 2012? Surveylab offers more than professionally designed surveys and easy-to-use reports. We have years of experience designing, implementing and managing customer feedback programmes (and more from past careers and partnering with specialist consultancies). Get in touch!

Occasionally, one of our potential clients has a similar reaction to Surveylab’s quote (which usually falls between £1,000 and £3,000) – they think

“but there are a plethora of services where I can do it myself for £50!”

Comparing apples and pears

This may be true, but the point is that YOU have to do it. As a survey company, we bring specific skills and experience to the project, much like a kitchen installer does when fitting a new kitchen. Keeping with my analogy, the kitchen installer can also help design the kitchen and provide after-sales care.

Positioned correctly, using an external survey company demonstrates to staff the importance in the survey and desire to ‘get it right’, with the added benefit of ensuring respondent anonymity. (sorry, couldn’t think of another kitchen analogy.)

Value for money?

This is the obvious question any buyer should be asking. In the case of an employee survey, if you have 150 staff, and outsourcing your survey costs £1,650, this equates to a cost of £11/employee to ask questions and get feedback from all your staff in a quick and efficient process. There are hidden costs (most significant will be your time) and perhaps the expense of printing and mailing letters, but £11 to engage and listen to each and every one of your staff is a very small cost that can deliver superb value.

For 1,000 staff, the cost is much less – in this example just £1.65. Expenses may be a little higher, but economies of scale should apply with everything except for any postage.

Ultimately, it’s what you do after the survey is closed that really determines whether your survey was value for money.